ESP8266 Community Forum

- Sun Feb 21, 2016 5:35 am #41578 As you know, it's tough getting below 50uA in practical applications. One other thing to bear in mind is the leakage current in electrolytic capacitors. What actual manufacturers part is your 100uF?

Personally I consider 50uA to be a reasonable average current draw representing a couple of years service on a 1AH cell so I'm curious as to why you would go to all this extra effort - don't get me wrong - it's a worthy effort

- Sun Feb 21, 2016 1:25 pm #41603 I am aware of capacitor leakage. Using all X5R ceramic capacitors on board.
According to: http://www.tme.eu/en/Document/73f8fd69a ... VX-X5R.pdf
their brand X5R's IR is 500MΩ / µF, so for 100uF the IR is 5MOhm. So all capacitors add together probably incur around 1uA leakage, on paper.

However, I believe the leakage is also correlated with the voltage applied. Putting 3.3v on 6.8v or 10v rated capacitor should yield less leakage than the nominal value.

My effort was originally driven by my lazyness - I want to stick my EPS sensors to the ceiling and don't have to worry about climb ladder to change battery for as long as possible.

Later after I learned quite a bit wrt. power supplies and power management, it became a technical curiosity - whether the state-of-the-art SMPS technology (commercially available, of course) could compete with the best linear regulation.

On my lazyness line, I have done other research, such as indoor ambient light energy harvesting. I explored a wide range of panels and boost adapters, once I collect more data and got more time, I will open another thread to share my experiences.

- Sun Feb 21, 2016 3:39 pm #41611 Good component choice... it's marvelous how far ceramic capacitors have come
BTW I've often wondered if energy could be harvested from low frequency pressure changes - the sort you get when doors are opened and closed in buildings.

- Sun Feb 21, 2016 7:26 pm #41619 I noticed several of my previous posts are gone, probably happened when this forum was upgraded.

So I opened a GitHub project for this exploration, moved files and documentations there for a higher availability:
https://github.com/Adam5Wu/Micro-power-SMPS

Yes, vibration energy can be harvested using piezoelectric generators (a reverse procedure of the piezoelectric buzzer on the birthday music card). It has very high voltage, but not much current.
Several chips claim to be able to harvest this kind of energy, like LTC3588-1, bq25570.

Haven't explored much along that line, though.